Artificial intelligence (AI) technology is rapidly being introduced and used in all industries as the core technology. Further, concerns about unexpected social issues are also emerging. Therefore, each country, and standard and international organizations, are developing and distributing guidelines to maximize the benefits of AI while minimizing risks and side effects. However, there are several hurdles for developers to use them in actual industrial fields such as ambiguity in terminologies, lack of concreteness according to domain, and non-specific requirements. Therefore, in this paper, approaches to address these problems are presented. If the recommendations or guidelines to be developed in the future refer to the proposed approaches, it would be a guideline for assuring AI trustworthiness that is more developer-friendly.

oct

Jeju Island, Korea, Republic of

https://ieeexplore.ieee.org/document/9953030/